Package ‘sExtinct’
February 20, 2015
Type Package
Title Calculates the historic date of extinction given a series of
sighting events
Version 1.1
Date 2012-10-24
Author Christopher Clements
Maintainer Christopher Clements <[email protected]>
Description This package combines several sighting based estimators of historical extinction, allowing them to be run simultaneously or individually. Code for this package was contributed by Ben Collen, Gene Hunt and Tracy Rout. Additional code was taken from McPherson & Myers (2009).
License GPL-2
Depends graphics, grDevices, stats, utils, lattice
NeedsCompilation no
Repository CRAN
Date/Publication 2013-10-31 08:06:49
R topics documented:
sExtinct-package .
example.data . . .
OLE . . . . . . . .
print.extmod . . . .
Robson1964 . . . .
run.all . . . . . . .
Solow1993.eq2 . .
Solow1993.eq2.fun
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Index
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
3
4
5
5
6
7
9
10
1
2
sExtinct-package
sExtinct-package
sExtinct
Description
sExtinct combines several sighting based estimators of historical extinction.
Details
Package:
Type:
Version:
Date:
License:
sExtinct
Package
1.1
2012-10-25
GPL-2 | GPL-3
This package brings together several sighting-based estimators of species loss. These estimators
use historical sighting events to calculate the probability of a species persistance. The esitimators
can be run simultaneously (using run.all), or individually (see help files). Care should be taken
if using low level functions (functions ending in .fun). New users should use intermediate level
functions (Robson1964, Solow1993.eq2, OLE) which correct data before running the estimators.
This is a work in progress, and additional estimators are sought. Please contact author (see below).
Author(s)
Package written by Christopher Clements <[email protected]>.
Code for functions was contributed by:
Ben Collen
Gene Hunt
Tracy Rout
Additional code take from the Appendix of McPherson & Myers (2009)
References
Burgman, M. A., Grimson, R. C., & Ferson, S. (1995). Inferring Threat from Scientific Collections.
Conservation Biology, 9(4), 923-928.
Roberts, D. L., & Solow, A. R. (2003). Flightless birds: When did the dodo become extinct? Nature,
426(6964), 245.
Robson, D. S., & Whitlock, J. H. (1964). Estimation of a Truncation Point. Biometrika, 31(1/2),
33-39.
Solow. (1993). Inferring extinction from sighting data. Ecology, 74(3), 962-964.
Solow. (2005). Inferring extinction from a sighting record. Mathematical Biosciences, 195(1),
47-55.
example.data
3
Strauss, D., & Sadler, P. M. (1989). Classical confidence intervals and Bayesian probability estimates for ends of local taxon ranges. Mathematical Geology, 21(4), 411-427.
See Also
run.all Robson1964 Solow1993.eq2 OLE
Examples
data(example.data)
run.all(example.data, 0.05, 2012, FALSE, FALSE)
example.data
Example data
Description
A fictional data set of historic sighting events.
Usage
data(example.data)
Format
A data frame with 8 observations on the following 2 variables.
years a numeric vector of the years sights have or have not occured at. Sorted from earliest at the
top to latest at the bottom. Restricted to one row per year.
sightings a numeric vector of the number of sighting events (if any) that have occured in that
year.
Examples
data(example.data)
example.data
4
OLE
OLE
Optimal linear estimation
Description
A function that calculates the predcted date of extinction using the optimal lienar estimation technique proposed by Roberts and Solow (2003) and Solow (2005). Differs from other techniques
in this package as it requires only data to produce a point estimate of the date of extinction (not
dependant on 1-alpha CI), and uses alpha to calculate the upper and lower 1-alpha CI.
Usage
OLE(sightingdata, alpha)
Arguments
sightingdata
A data.frame with two columns, the first containing the year at which a sighting
has (or as not) occured, the second a column of the number of sightings that
have occured in each year. Column names are not restricted. Data msut be
ranked from oldest at the top, to most recent at the bottom, and restricted to a
single row per year. See example.data.
alpha
A 1-alpha confidence interval is calculated for a given value of alpha, typically
set as 0.05 (i.e. a 95% CI). OLE produces an upper and lower confidence interval
estimate, the lower obviously bound by the last sighting event. See Clements et
al. for a discussion on the use of 95% CI’s produced by OLE.
Author(s)
Christopher Clements
References
Clements, C. F., Worsfold, N., Warren, P., Collen, B., Blackburn, T., Clark, N., & Petchey, O. L.
Experimentally testing an extinction estimator: Solows Optimal Linear Estimation model. Journal
of Animal Ecology, In press.
Roberts, D. L., & Solow, A. R. (2003). Flightless birds: When did the dodo become extinct? Nature,
426(6964), 245.
Solow. (2005). Inferring extinction from a sighting record. Mathematical Biosciences, 195(1),
47-55.
Examples
data(example.data)
OLE.fun(example.data, alpha=0.05)
print.extmod
5
print.extmod
Result style
Description
Defines how extinction predictor results are displayed.
Usage
## S3 method for class 'extmod'
print(x,...)
## S3 method for class 'simpextmod'
print(x,...)
Arguments
x
results from the estimator
...
...
Robson1964
Robson1964, Strauss89
Description
Sighting based estimators of extinction that will calculate p values for persistance at each year
after the last sighting event of a species. Unlike Burgman, Solow1993.eq2, Solow2005.eq7 these
predictors do not need a test.year.
Usage
Robson1964(sightingdata, alpha, data.out)
Strauss89(sightingdata, alpha, data.out)
Arguments
sightingdata
A data.frame with two columns, the first containing the year at which a sighting
has (or as not) occured, the second a column of the number of sightings that
have occured in each year. Column names are not restricted. Data msut be
ranked from oldest at the top, to most recent at the bottom, and restricted to a
single row per year. See example.data.
alpha
A 1-alpha confidence interval is calculated for a given value of alpha, typically
set as 0.05 (i.e. a 95%CI).
data.out
A year which chance of persistance is calculated up until, typically set as the
current year (i.e. test whether the species is currently predicted to be extant or
extinct).
6
run.all
Author(s)
Christopher Clements
References
Robson, D. S., & Whitlock, J. H. (1964). Estimation of a Truncation Point. Biometrika, 31(1/2),
33-39.
Strauss, D., & Sadler, P. M. (1989). Classical confidence intervals and Bayesian probability estimates for ends of local taxon ranges. Mathematical Geology, 21(4), 411-427.
Examples
data(example.data)
Robson1964(example.data, alpha=0.05, data.out=FALSE)
run.all
Run all extinction estimators
Description
A function that runs all the sighting based extinction predictors within this package. The function
can produce diagnostic plots (utilising the package lattice) and produce simple, and detailed data
outputs.
Usage
run.all(sightingdata, alpha, test.year, data.out, plot)
Arguments
sightingdata
A data.frame with two columns, the first containing the year at which a sighting
has (or as not) occured, the second a column of the number of sightings that
have occured in each year. Column names are not restricted. Data msut be
ranked from oldest at the top, to most recent at the bottom, and restricted to a
single row per year. See example.data
alpha
A 1-alpha confidence interval is calculated for a given value of alpha, typically
set as 0.05 (i.e. a 95%CI).
test.year
A year which chance of persistance is calculated up until, typically set as the
current year (i.e. test whether the species is currently predicted to be extant or
extinct).
data.out
A TRUE/FALSE argument where TRUE produces the predicted chance of persistance at each time step (where applicable), and FALSE gives a simple output
with the predicted date of extinction only.
plot
A TRUE/FALSE argument where TRUE produces a diagnostic plot showing the
probability of persistance estimated by each method, where applicable, along
with the alpha level set.
Solow1993.eq2
7
Value
If data.out=TRUE, a data.frame of probability of persistance at each time step, for each method,
except OLE which produces a point estimate of extinction only.
If data.out=FALSE, a data.frame of the predicted date of extinction only, produced by each method.
NOTE: upper and lower CIs are only calculated OLE, and will be filled in with NAs for the other
estimators.
Author(s)
Christopher Clements
References
Burgman, M. A., Grimson, R. C., & Ferson, S. (1995). Inferring Threat from Scientific Collections.
Conservation Biology, 9(4), 923-928.
Roberts, D. L., & Solow, A. R. (2003). Flightless birds: When did the dodo become extinct? Nature,
426(6964), 245.
Robson, D. S., & Whitlock, J. H. (1964). Estimation of a Truncation Point. Biometrika, 31(1/2),
33-39.
Solow. (1993). Inferring extinction from sighting data. Ecology, 74(3), 962-964.
Solow. (2005). Inferring extinction from a sighting record. Mathematical Biosciences, 195(1),
47-55.
Strauss, D., & Sadler, P. M. (1989). Classical confidence intervals and Bayesian probability estimates for ends of local taxon ranges. Mathematical Geology, 21(4), 411-427.
Examples
data(example.data)
run.all(example.data, alpha=0.05, test.year=2012, data.out=FALSE, plot=FALSE)
Solow1993.eq2
Intermediate level functions to calculate Burgman, Solow1993.eq2,
Solow2005.eq7
Description
A collection of functions requiring, in addition to the normal intermediate level inputs of data,
alpha and data.out, a test.year value. The test.year (typically set as the current year) is the date that
p values of persistance are calculated up to.
Usage
Burgman(sightingdata, alpha, test.year, data.out)
Solow1993.eq2(sightingdata, alpha, test.year, data.out)
Solow2005.eq7(sightingdata, alpha, test.year, data.out)
8
Solow1993.eq2
Arguments
sightingdata
A data.frame with two columns, the first containing the year at which a sighting
has (or as not) occured, the second a column of the number of sightings that have
occured in each year. Can be gaps in the data (which are filled in by the function
as 0 sightings within that year). Column names are not restricted. Data must be
ranked from oldest at the top, to most recent at the bottom, and restricted to a
single row per year. See example.data.
alpha
A 1-alpha confidence interval is calculated for a given value of alpha, typically
set as 0.05 (i.e. a 95%CI).
test.year
A year which chance of persistance is calculated up until, typically set as the
current year (i.e. test whether the species is currently predicted to be extant or
extinct).
data.out
A TRUE/FALSE argument where TRUE produces the predicted chance of persistance at each time step (where applicable), and FALSE gives a simple output
with the predicted date of extinction only.
Note
This is a work in progress, to submit a new (to this package) sighting based method for inferring
prediction please contact the author (details can be found on www.chrisclementsresearch.co.uk)
Author(s)
Christopher Clements
References
Burgman, M. A., Grimson, R. C., and Ferson, S. (1995). Inferring Threat from Scientific Collections. Conservation Biology, 9(4), 923-928. Solow (1993). Inferring extinction from sighting data.
Ecology, 74(3), 962-964. Solow (2005). Inferring extinction from a sighting record. Mathematical
Biosciences, 195(1), 47-55.
See Also
Solow1993.eq2.fun
Examples
data(example.data)
Burgman(example.data, alpha=0.05, test.year=2012, data.out=FALSE)
Solow1993.eq2.fun
9
Solow1993.eq2.fun
Low level functions for Burgman, Solow1993.eq2, Solow2005.eq7
Description
Low level functions on which intermediate level functions are based. Included here for advanced
users only. Functions take only a data.frame with two columns, the first containing the year at which
a sighting has (or as not) occured, the second a column of the number of sightings that have occured
in each year. There must not be missing values in the data frames (i.e. years where sightings=0
must also be included). Functions will calculate the p-value of persistance for the final year in the
data frame. To calculate the p-value for the present day the data.frame must contain all the years
since the last sighting event with 0 sightings (this is automatically added by the intermediate level
functions).
Usage
Burgman.fun(dd)
Solow1993.eq2.fun(dd)
Solow2005.eq7.fun(dd)
Arguments
dd
A data.frame with two columns, the first containing the year at which a sighting
has (or as not) occured, the second a column of the number of sightings that have
occured in each year. Gaps in sighting history must be filled, see description
above.Column names are not restricted. Data must be ranked from oldest at the
top, to most recent at the bottom, and restricted to a single row per year.
Author(s)
Christopher Clements
References
Burgman, M. A., Grimson, R. C., & Ferson, S. (1995). Inferring Threat from Scientific Collections.
Conservation Biology, 9(4), 923-928. Solow. (1993). Inferring extinction from sighting data.
Ecology, 74(3), 962-964. Solow. (2005). Inferring extinction from a sighting record. Mathematical
Biosciences, 195(1), 47-55.
Examples
years <- c(1950:2012)
sightings <- c(1,0,3,4,0,1,0,1,(c(rep(0, 55))))
example.data.for.simple.functions<-data.frame(years, sightings)
Solow2005.eq7.fun(example.data.for.simple.functions)
Index
∗Topic datasets
example.data, 3
∗Topic package
sExtinct-package, 2
Burgman (Solow1993.eq2), 7
Burgman.fun (Solow1993.eq2.fun), 9
example.data, 3, 6
OLE, 2, 3, 4
print.extmod, 5
print.simpextmod (print.extmod), 5
Robson1964, 2, 3, 5
run.all, 2, 3, 6
sExtinct (sExtinct-package), 2
sExtinct-package, 2
Solow1993.eq2, 2, 3, 7
Solow1993.eq2.fun, 8, 9
Solow2005.eq7 (Solow1993.eq2), 7
Solow2005.eq7.fun (Solow1993.eq2.fun), 9
Strauss89 (Robson1964), 5
Strauss89full.fun (Robson1964), 5
10